A novel estimation scheme that combines Bayesian and lower bound estimating radio frequency identification tag population size is proposed. The developed methodology is based on the fusion between the Bayesian and lower bound estimating techniques. It turns out that the fusion rule is built up thanks to an existing linear relationship between the cited techniques. Simulation resultsshow that the developed technique significantly improves the accuracy of the estimating tag quantity and presents less estimation error. Also, the resulting advanced dynamic framed slotted ALOHA protocol considerably improves the performance and efficiency of the radio frequency identification anti-collision compared with the most recent protocols using others estimating methods.
KEYWORDSBayesian estimation, lower bound, RFID, slotted ALOHA
INTRODUCTIONRadio frequency identification (RFID) technology is becoming an integral part in a wide spectrum of applications, such as access control, distribution, medicine, and RFID-embedded cards. However, despite the significant increasing of the density of available tags in the readers' interrogation zones, the tag collision problem remains a serious issue that must be addressed seriously before any deployment. The collision problem occurs when simultaneous tags share the same communication channel with the interrogating reader. As a result, the reader will not be able to get the correct identity from each tag. Recently, several state-of-the-art anti-collision tag identification algorithms have been proposed in the literature to mitigate the low tag identification efficiency and performance.The anti-collision algorithm can be divided into 2 main categories. 1 The first category concerns the anti-collision techniques that are processed in medium access control (MAC) layer. In this category, most of the RFID reader uses dynamic framed slotted ALOHA (DFSA)-based probabilistic algorithms. 2-7 The DFSA protocols dynamically adjust the frame size in the successive rounds according to the estimating tags quantity. In fact, DFSA anti-collision protocols can be divided into 2 key parts. The first one deals on how to estimate the tag quantity during an inventory process. The second part focuses on how to dynamically adjust frame size, according to the estimated number of tags, to achieve maximum throughput. Knowing that, the tag quantity to be read is usually unknown, and the maximum throughput of the system is achieved when the frame length is equal to the number of tags of the system. 8 Therefore, it is utmost important to use an estimation method of high accuracy in order to not affect the reading performance of DFSA algorithms.In this context, numerous methods have been proposed in the literature. Among conventional methods (process the problem in MAC layer), the basic ones are idle slot estimate, 9 lower bound estimate, 10 and Schoute estimate. 11 So far, Int J Commun Syst. 2018;31:e3723.wileyonlinelibrary.com/journal/dac